Comparison of hydrographic and altimeter based estimates of sea level height variability in the Atlantic Ocean

Abstract

Abstract Our ability to understand the means by which mass and heat are exchanged between the tropics and subtropics is seriously compromised when using only sea level data because the exchange processes span a wide range of variability across the different dynamical regimes in our domain. Expendable bathythermograph (XBT) profiles and TOPEX/Poseidon (T/P) altimeter data are compared to temperature anomalies (TA) and to dynamic height anomalies (DHA) for the period 1993 through 1997 to determine how much can be inferred about the internal field of mass from sea level changes. Our focus is on the annual cycle along two well-sampled XBT sections on the western and eastern sides of the Atlantic Ocean from 10°S to 40°N. XBT profiles were matched (time/location) to Sea height anomalies (SHA) derived from T/P data, converted into DHA using TS relationships and then binned monthly into 2° of latitude by 4° of longitude boxes. The vertical mass distribution cannot always be inferred from SHA alone, unless there is a strong relationship between SHA and DHA and an understanding of the details of how temperature variability affects DHA. These relationships can be problematic if SHA are small. This occurs in zones of transition in the vicinity of troughs where small fluctuations in SHA belie the true nature of water column variability. These areas separate the mid-latitudes where surface buoyancy fluxes dominate from those in the equatorial region where ocean dynamics cause thermocline effects that dominate the forcing of sea level. Thus, the variability of SHA in transition regions tends to be small because both surface and thermocline variability may be significant but compensating in nature. This emphasizes how important direct observations (in situ data) can be in interpreting SHA correctly. Strong relationships between SHA and DHA occur at locations where more than half of the SHA variance in the annual cycle is due to DHA variability (approximately 30% of the positions along the two XBT sections). These relationships between SHA and DHA for residual variability (obtained by removing the annual cycle) are weak. The exceptions are in two areas of large sea height variability in the western basin where there is significant interannual variability. The first is in the tropics in the vicinity of the tropical gyre trough near 50°W 8°N. The second is in the Gulf Stream near 70°W 38°N. An analysis of Panulirus data at (32.2°N, 64.5°W) suggests that in situ data may be needed down to at least 1000m where interannual variability accounts for about 40% of the SHA variance.